From mid-century dreams to digital dilemmas: unravelling the evolution of AI
Learn about the 70 year history of AI in this Thought Piece by the NSSN’s Machine Learning and Data Engineer, Dr Ayu Saraswati.
Artificial intelligence (AI) has seemingly become ubiquitous and prolific overnight.
However, its origins date back to mid-century research when the seminal event Dartmouth Conference in 1956 took place.
AI emerged as a visionary pursuit aimed at replicating human cognitive capabilities in machines.
Today, when we talk about AI, we are often referring to a specific type of machine learning algorithm known as Artificial Neural Networks (ANN).
The AI term, however, is a broad one and encompasses a wide range of technologies and techniques.
Inspired by the brain
The concepts of AI were dreamt up in the 1950s and 1960s but are only now becoming a reality. Credit: Shutterstock
At its core, ANN draws inspiration from the human brain.
Artificial perceptron, the building blocks of neural networks, mimic the behaviour of biological neurons.
Over time, different network architectures have emerged, including convolutional neural networks (CNNs) and transformers.
These architectures can be combined in various ways, leading to innovations like generative AI.
Interestingly, we’ve come full circle in a convergence of AI and neuroscience.
Recently, researchers have grown brain cells in the lab and taught them to play the game Pong.
This experiment shows how closely AI and biology can intersect.
Why now?
If the concepts of AI were dreamt up in the 1950s and 1960s, why is it only now that they’re becoming a reality?
The answer lies in two key developments: the proliferation of digital data and advancements in hardware technology.
The exponential growth of the internet and social media platforms has led to an explosion in data.
This big data surge is both a problem that AI can solve and a requirement for it to function effectively.
On the hardware front, the development of Graphics Processing Units (GPUs) has provided the computational power necessary to process large amounts of data and run complex AI algorithms.
This synergy between data availability and computational prowess has propelled AI into the forefront of innovation across diverse sectors.
Dr Ayu Saraswati. Credit: NSSN
AI has the potential to revolutionise many fields.
In scientific discovery, it can aid in drug discovery and provide new imaging opportunities in medicine and astronomy.
It can enhance inclusivity and accessibility, with technologies like automated captions for the hearing impaired and emotional recognition for neurodivergent individuals.
Moreover, AI enables novel creative possibilities, from digitally rejuvenating beloved characters to ensuring the safety of performers through virtual stunt doubles.
Ethical consequences
However, AI’s potential to do good and its accelerated proliferation is not without ethical and societal implications.
The abundance of data raises concerns about privacy, security, and the equitable distribution of AI benefits.
Foundational AI models often come from tech giants like Google, Meta, and Microsoft, who possess both data and compute resources.
This could lead to a monopoly and widen the gap between the haves and have-nots.
AI can also contribute to the spread of misinformation at an unprecedented volume and speed.
It can be used for the non-consensual use of other people’s images, such as in revenge porn.
And because AI learns from historical data, it can amplify existing biases.
Navigating these digital dilemmas necessitates a multifaceted and interdisciplinary approach to AI governance.
Addressing AI bias and promoting diversity, equity, and inclusion in AI research and development are imperative for fostering ethical AI deployment.
Furthermore, fostering collaboration among technologists, policymakers, ethicists, and civil society stakeholders is essential for shaping a future where AI serves as a catalyst for social good and human welfare.
The evolution of AI from its mid-century origins to its contemporary complexities reflects a convergence of technological ingenuity and societal needs.
AI potential for good or harm lies in our choices.
As we continue to harness its power, we must remember that AI is a tool, not a panacea.
It can help us solve complex problems and improve lives, but it can also exacerbate existing inequalities and create new ones.
Therefore, it is incumbent upon us to guide its development and application with wisdom, foresight, and a deep commitment to ethical principles.
The NSW Smart Sensing Network (NSSN) is uniquely positioned to foster ethical and responsible AI innovation through its collaborative model that engages researchers, government entities, and industry stakeholders. By harnessing this collective expertise, the NSSN can contribute to the development of AI-driven sensor systems that benefits everyone. For more information contact Dr Ayu Saraswati.